Effects of sound-to-spelling friends and enemies in children's auditory rhyme decisions

Eleni Lipourli

doi:10.18844/prosoc.v7i1.4902

Abstract

An experiment is reported, conducted on children of 9 and 11 years of age, that examined the possible effect of the relative number of sound-to-spelling friends and enemies the first word of a rhyming pair has on performance in the auditory rhyme judgment task. Friends are rhyming words with the same spelling of the rime (e.g., stream-dream and scheme-theme) and enemies are rhyming words with dissimilar spellings of the rimes (e.g., stream-theme and scheme-dream). It was predicted, following the Loosemore, Brown & Watson (1991) connectionist model of spelling, that the first word presented should activate the representations of its friends and inhibit the representations of its enemies and so the orthographic facilitation effect should be larger for those cases where the first word a rhyming pair has more friends than more enemies. The experiment reported in this paper found precisely the opposite pattern of results: the orthographic facilitation effect was only reliable for those cases when the first word had more enemies than friends. So it would appear to be the case that the degree of the activation of friends and the inhibition of enemies varies according to how many words are friends and how many words are enemies.  Keywords: rhyme judgment task; friends and enemies; orthographic facilitation effect; children.

Highlights

Spelling-to-sound relationships may be described in different ways
The division into ‘regular’ and ‘exception’ words for reading has been complicated by a number of issues, including: (i) what is the effective size of the units for which correspondence rules should operate; (ii) whether regularity should be determined by the number of words or the absolute frequency of the words; and (iii) how flexible should rules be in terms of allowing the operation of context-sensitivity constraints (e.g., A is usually pronounced as /a/, except when it follows W, when it is often pronounced as /o/, as in was, wasp and wash)
The mean reaction time of all the twenty-four 9-year-olds was 1322 msec and the mean reaction time of all the twenty-two 11-year-olds was 1325 msec, and this difference was not significant [t(44) = 0.05]. This experiment was designed to see if children of 9 and 11 years of age would show effects of sound-to-spelling friends and enemies in the auditory rhyme judgment task

Summary

Introduction

Spelling-to-sound relationships may be described in different ways. Traditional symbolic rule-based methods have stressed the concept of regularity and have partitioned words into those with ‘regular’ correspondences (i.e., whose pronunciations are correctly determined by a set of spelling-to-sound translation rules) and those with ‘irregular’ or ‘exception’ correspondences (i.e., whose pronunciations are different from those given by the rules). If the word “steam” was presented first, it should be expected to activate its many friends (all of which, by definition, have a similar spelling) If it was followed, in the auditory rhyme decision task, with one of these friends, it is quite reasonable to expect a large orthographic facilitation effect, compared to when one of its fewer enemies (i.e., a word with a different spelling pattern) is presented. If the Loosemore et al model of spelling activation can be applied to performance in the auditory rhyme decision task, it is predicted that the orthographic facilitation effect should be larger for those rhyming word pairs where the first word has more friends than where it has more enemies This prediction will be tested in both 9- and 11-year-old children. The experiment manipulated two variables: the first word of each rhyming pair either had more friends (e.g., “steam”) or more enemies (e.g., “scheme”), and was followed by either a spelled word (i.e., a friend of the first word) or a dissimilarly spelled word (i.e., an enemy of the first word)

Method

Results

Discussion